Speak, Memory: An Archaeology Of Books Known To Chatgpt/gpt-4

Kent K. Chang, Mackenzie Cramer, Sandeep Soni, David Bamman. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing 2023 – 28 citations

[Paper]    
Reinforcement Learning Training Techniques Model Architecture GPT

In this work, we carry out a data archaeology to infer books that are known to ChatGPT and GPT-4 using a name cloze membership inference query. We find that OpenAI models have memorized a wide collection of copyrighted materials, and that the degree of memorization is tied to the frequency with which passages of those books appear on the web. The ability of these models to memorize an unknown set of books complicates assessments of measurement validity for cultural analytics by contaminating test data; we show that models perform much better on memorized books than on non-memorized books for downstream tasks. We argue that this supports a case for open models whose training data is known.

Similar Work